| Message ID | 1584502378-12609-2-git-send-email-iamjoonsoo.kim@lge.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | integrate classzone_idx and high_zoneidx | expand |
On Wed, 18 Mar 2020, js1304@gmail.com wrote: > From: Joonsoo Kim <iamjoonsoo.kim@lge.com> > > Currently, we use the zone index of preferred_zone which represents > the best matching zone for allocation, as classzone_idx. It has a problem > on NUMA system with ZONE_MOVABLE. > Hi Joonsoo, More specifically, it has a problem on NUMA systems when the lowmem reserve protection exists for some zones on a node that do not exist on other nodes, right? In other words, to make sure I understand correctly, if your node 1 had a ZONE_MOVABLE than this would not have happened. If that's true, it might be helpful to call out that ZONE_MOVABLE itself is not necessarily a problem, but a system where one node has ZONE_NORMAL and ZONE_MOVABLE and another only has ZONE_NORMAL is the problem. > In NUMA system, it can be possible that each node has different populated > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and > node 1 could have only NORMAL zone. In this setup, allocation request > initiated on node 0 and the one on node 1 would have different > classzone_idx, 3 and 2, respectively, since their preferred_zones are > different. If they are handled by only their own node, there is no problem. I'd say "If the allocation is local" rather than "If they are handled by only their own node". > However, if they are somtimes handled by the remote node due to memory > shortage, the problem would happen. > > In the following setup, allocation initiated on node 1 will have some > precedence than allocation initiated on node 0 when former allocation is > processed on node 0 due to not enough memory on node 1. They will have > different lowmem reserve due to their different classzone_idx thus > an watermark bars are also different. > > root@ubuntu:/sys/devices/system/memory# cat /proc/zoneinfo > Node 0, zone DMA > per-node stats > ... > pages free 3965 > min 5 > low 8 > high 11 > spanned 4095 > present 3998 > managed 3977 > protection: (0, 2961, 4928, 5440) > ... > Node 0, zone DMA32 > pages free 757955 > min 1129 > low 1887 > high 2645 > spanned 1044480 > present 782303 > managed 758116 > protection: (0, 0, 1967, 2479) > ... > Node 0, zone Normal > pages free 459806 > min 750 > low 1253 > high 1756 > spanned 524288 > present 524288 > managed 503620 > protection: (0, 0, 0, 4096) > ... > Node 0, zone Movable > pages free 130759 > min 195 > low 326 > high 457 > spanned 1966079 > present 131072 > managed 131072 > protection: (0, 0, 0, 0) > ... > Node 1, zone DMA > pages free 0 > min 0 > low 0 > high 0 > spanned 0 > present 0 > managed 0 > protection: (0, 0, 1006, 1006) > Node 1, zone DMA32 > pages free 0 > min 0 > low 0 > high 0 > spanned 0 > present 0 > managed 0 > protection: (0, 0, 1006, 1006) > Node 1, zone Normal > per-node stats > ... > pages free 233277 > min 383 > low 640 > high 897 > spanned 262144 > present 262144 > managed 257744 > protection: (0, 0, 0, 0) > ... > Node 1, zone Movable > pages free 0 > min 0 > low 0 > high 0 > spanned 262144 > present 0 > managed 0 > protection: (0, 0, 0, 0) > > min watermark for NORMAL zone on node 0 > allocation initiated on node 0: 750 + 4096 = 4846 > allocation initiated on node 1: 750 + 0 = 750 > > This watermark difference could cause too many numa_miss allocation > in some situation and then performance could be downgraded. > > Recently, there was a regression report about this problem on CMA patches > since CMA memory are placed in ZONE_MOVABLE by those patches. I checked > that problem is disappeared with this fix that uses high_zoneidx > for classzone_idx. > > http://lkml.kernel.org/r/20180102063528.GG30397@yexl-desktop > > Using high_zoneidx for classzone_idx is more consistent way than previous > approach because system's memory layout doesn't affect anything to it. > With this patch, both classzone_idx on above example will be 3 so will > have the same min watermark. > > allocation initiated on node 0: 750 + 4096 = 4846 > allocation initiated on node 1: 750 + 4096 = 4846 > Alternatively, I assume that this could also be fixed by changing the value of the lowmem protection on the node without managed pages in the upper zone to be the max protection from the lowest zones? In your example, node 1 ZONE_NORMAL would then be (0, 0, 0, 4096). > One could wonder if there is a side effect that allocation initiated on > node 1 will use higher bar when allocation is handled on node 1 since > classzone_idx could be higher than before. It will not happen because > the zone without managed page doesn't contributes lowmem_reserve at all. > > Reported-by: Ye Xiaolong <xiaolong.ye@intel.com> > Tested-by: Ye Xiaolong <xiaolong.ye@intel.com> > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com> Curious: is this only an issue when vm.numa_zonelist_order is set to Node? > --- > mm/internal.h | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/mm/internal.h b/mm/internal.h > index c39c895..aebaa33 100644 > --- a/mm/internal.h > +++ b/mm/internal.h > @@ -119,7 +119,7 @@ struct alloc_context { > bool spread_dirty_pages; > }; > > -#define ac_classzone_idx(ac) zonelist_zone_idx(ac->preferred_zoneref) > +#define ac_classzone_idx(ac) (ac->high_zoneidx) > > /* > * Locate the struct page for both the matching buddy in our > -- > 2.7.4 > > >
2020년 3월 19일 (목) 오전 6:29, David Rientjes <rientjes@google.com>님이 작성: > > On Wed, 18 Mar 2020, js1304@gmail.com wrote: > > > From: Joonsoo Kim <iamjoonsoo.kim@lge.com> > > > > Currently, we use the zone index of preferred_zone which represents > > the best matching zone for allocation, as classzone_idx. It has a problem > > on NUMA system with ZONE_MOVABLE. > > > > Hi Joonsoo, Hello, David. > More specifically, it has a problem on NUMA systems when the lowmem > reserve protection exists for some zones on a node that do not exist on > other nodes, right? Right. > In other words, to make sure I understand correctly, if your node 1 had a > ZONE_MOVABLE than this would not have happened. If that's true, it might > be helpful to call out that ZONE_MOVABLE itself is not necessarily a > problem, but a system where one node has ZONE_NORMAL and ZONE_MOVABLE and > another only has ZONE_NORMAL is the problem. Okay. I will try to re-write the commit message as you suggested. > > In NUMA system, it can be possible that each node has different populated > > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and > > node 1 could have only NORMAL zone. In this setup, allocation request > > initiated on node 0 and the one on node 1 would have different > > classzone_idx, 3 and 2, respectively, since their preferred_zones are > > different. If they are handled by only their own node, there is no problem. > > I'd say "If the allocation is local" rather than "If they are handled by > only their own node". I will replace it with yours. Thanks for correcting. > > However, if they are somtimes handled by the remote node due to memory > > shortage, the problem would happen. > > > > In the following setup, allocation initiated on node 1 will have some > > precedence than allocation initiated on node 0 when former allocation is > > processed on node 0 due to not enough memory on node 1. They will have > > different lowmem reserve due to their different classzone_idx thus > > an watermark bars are also different. > > > > root@ubuntu:/sys/devices/system/memory# cat /proc/zoneinfo > > Node 0, zone DMA > > per-node stats > > ... > > pages free 3965 > > min 5 > > low 8 > > high 11 > > spanned 4095 > > present 3998 > > managed 3977 > > protection: (0, 2961, 4928, 5440) > > ... > > Node 0, zone DMA32 > > pages free 757955 > > min 1129 > > low 1887 > > high 2645 > > spanned 1044480 > > present 782303 > > managed 758116 > > protection: (0, 0, 1967, 2479) > > ... > > Node 0, zone Normal > > pages free 459806 > > min 750 > > low 1253 > > high 1756 > > spanned 524288 > > present 524288 > > managed 503620 > > protection: (0, 0, 0, 4096) > > ... > > Node 0, zone Movable > > pages free 130759 > > min 195 > > low 326 > > high 457 > > spanned 1966079 > > present 131072 > > managed 131072 > > protection: (0, 0, 0, 0) > > ... > > Node 1, zone DMA > > pages free 0 > > min 0 > > low 0 > > high 0 > > spanned 0 > > present 0 > > managed 0 > > protection: (0, 0, 1006, 1006) > > Node 1, zone DMA32 > > pages free 0 > > min 0 > > low 0 > > high 0 > > spanned 0 > > present 0 > > managed 0 > > protection: (0, 0, 1006, 1006) > > Node 1, zone Normal > > per-node stats > > ... > > pages free 233277 > > min 383 > > low 640 > > high 897 > > spanned 262144 > > present 262144 > > managed 257744 > > protection: (0, 0, 0, 0) > > ... > > Node 1, zone Movable > > pages free 0 > > min 0 > > low 0 > > high 0 > > spanned 262144 > > present 0 > > managed 0 > > protection: (0, 0, 0, 0) > > > > min watermark for NORMAL zone on node 0 > > allocation initiated on node 0: 750 + 4096 = 4846 > > allocation initiated on node 1: 750 + 0 = 750 > > > > This watermark difference could cause too many numa_miss allocation > > in some situation and then performance could be downgraded. > > > > Recently, there was a regression report about this problem on CMA patches > > since CMA memory are placed in ZONE_MOVABLE by those patches. I checked > > that problem is disappeared with this fix that uses high_zoneidx > > for classzone_idx. > > > > http://lkml.kernel.org/r/20180102063528.GG30397@yexl-desktop > > > > Using high_zoneidx for classzone_idx is more consistent way than previous > > approach because system's memory layout doesn't affect anything to it. > > With this patch, both classzone_idx on above example will be 3 so will > > have the same min watermark. > > > > allocation initiated on node 0: 750 + 4096 = 4846 > > allocation initiated on node 1: 750 + 4096 = 4846 > > > > Alternatively, I assume that this could also be fixed by changing the > value of the lowmem protection on the node without managed pages in the > upper zone to be the max protection from the lowest zones? In your > example, node 1 ZONE_NORMAL would then be (0, 0, 0, 4096). No, if lowmem_reserve of node 0 ZONE_NORMAL is (0, 0, 4096, 4096), min watermark of the allocation initiated on node 1 is 750 + 4096(classzone_idx 2) when allocation is tried on node 0 ZONE_NORMAL and issue would be gone. So, I think that it cannot be fixed by your alternative. > > One could wonder if there is a side effect that allocation initiated on > > node 1 will use higher bar when allocation is handled on node 1 since > > classzone_idx could be higher than before. It will not happen because > > the zone without managed page doesn't contributes lowmem_reserve at all. > > > > Reported-by: Ye Xiaolong <xiaolong.ye@intel.com> > > Tested-by: Ye Xiaolong <xiaolong.ye@intel.com> > > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com> > > Curious: is this only an issue when vm.numa_zonelist_order is set to Node? Do you mean "/proc/sys/vm/numa_zonelist_order"? It looks like it's gone now. Thanks.
On 3/19/20 9:57 AM, Joonsoo Kim wrote: >> Curious: is this only an issue when vm.numa_zonelist_order is set to Node? > > Do you mean "/proc/sys/vm/numa_zonelist_order"? It looks like it's gone now. > > Thanks. Yes it's gone now, but indeed, AFAIU on older kernels with zone order instead of node order, this problem wouldn't manifest. This was in my reply to v1, 2 years ago :) So to summarize; - ac->high_zoneidx is computed via the arcane gfp_zone(gfp_mask) and represents the highest zone the allocation can use - classzone_idx was supposed to be the highest zone that the allocation can use, that is actually available in the system. Somehow that became the highest zone that is available on the preferred node (in the default node-order zonelist), which causes the watermark inconsistencies you mention.
On 3/18/20 4:32 AM, js1304@gmail.com wrote: > From: Joonsoo Kim <iamjoonsoo.kim@lge.com> > > Currently, we use the zone index of preferred_zone which represents > the best matching zone for allocation, as classzone_idx. It has a problem > on NUMA system with ZONE_MOVABLE. > > In NUMA system, it can be possible that each node has different populated > zones. For example, node 0 could have DMA/DMA32/NORMAL/MOVABLE zone and > node 1 could have only NORMAL zone. In this setup, allocation request > initiated on node 0 and the one on node 1 would have different > classzone_idx, 3 and 2, respectively, since their preferred_zones are > different. If they are handled by only their own node, there is no problem. > However, if they are somtimes handled by the remote node due to memory > shortage, the problem would happen. > > In the following setup, allocation initiated on node 1 will have some > precedence than allocation initiated on node 0 when former allocation is > processed on node 0 due to not enough memory on node 1. They will have > different lowmem reserve due to their different classzone_idx thus > an watermark bars are also different. ... > Reported-by: Ye Xiaolong <xiaolong.ye@intel.com> > Tested-by: Ye Xiaolong <xiaolong.ye@intel.com> > Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com> With the clarifications that David requested, Acked-by: Vlastimil Babka <vbabka@suse.cz> Thanks! > --- > mm/internal.h | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/mm/internal.h b/mm/internal.h > index c39c895..aebaa33 100644 > --- a/mm/internal.h > +++ b/mm/internal.h > @@ -119,7 +119,7 @@ struct alloc_context { > bool spread_dirty_pages; > }; > > -#define ac_classzone_idx(ac) zonelist_zone_idx(ac->preferred_zoneref) > +#define ac_classzone_idx(ac) (ac->high_zoneidx) > > /* > * Locate the struct page for both the matching buddy in our >
2020년 3월 19일 (목) 오후 9:21, Vlastimil Babka <vbabka@suse.cz>님이 작성: > > On 3/19/20 9:57 AM, Joonsoo Kim wrote: > >> Curious: is this only an issue when vm.numa_zonelist_order is set to Node? > > > > Do you mean "/proc/sys/vm/numa_zonelist_order"? It looks like it's gone now. > > > > Thanks. > > Yes it's gone now, but indeed, AFAIU on older kernels with zone order instead of > node order, this problem wouldn't manifest. Yes. In this case, preferred_zone of an allocation is the populated highest zone among all nodes so classzone_idx will the same. Thanks.
diff --git a/mm/internal.h b/mm/internal.h index c39c895..aebaa33 100644 --- a/mm/internal.h +++ b/mm/internal.h @@ -119,7 +119,7 @@ struct alloc_context { bool spread_dirty_pages; }; -#define ac_classzone_idx(ac) zonelist_zone_idx(ac->preferred_zoneref) +#define ac_classzone_idx(ac) (ac->high_zoneidx) /* * Locate the struct page for both the matching buddy in our